Montar y desmontar

Rationale and objectives: Although onset of drug use during adolescence appears to increase long-term vulnerability to drug dependence in humans, relatively little is known about extinction and reinstatement of drug-seeking after periadolescent onset of drug self-administration in laboratory animals. Furthermore, although cue-induced reinstatement of cocaine-seeking

increases progressively during abstinence from cocaine self-administration in adult subjects, this “incubation of cocaine craving” remains unexplored after adolescent drug intake in animal models.

Methods: We allowed periadolescent (postnatal day 35; PND35 at start) and adult (PND83-95 at start) male Wistar rats to self-administer cocaine (0.36 mg/kg/infusion) in 2-h daily sessions on a fixed ratio 1 (FR1) schedule of reinforcement over 14 days. Then we compared extinction and cue- or cocaine priming-induced reinstatement (10 mg/kg cocaine, i.p.) of cocaine-seeking in both age groups after 30 days of abstinence in home cages. In separate cohorts, we tested for time-dependent increases in cue-induced reinstatement over approximately 1, 14, 30 or 60 days of abstinence in both age groups.

Results: Adolescent and adult rats self-administered similar amounts of cocaine. Subsequent cue-induced reinstatement was lower in the adolescent-onset group after a 30-day abstinence period, but cocaine priming-induced reinstatement did not differ across ages. Also extinction responding and time-dependent increases in cue-induced reinstatement (incubation) were less pronounced in rats that took cocaine as adolescents, compared with adults.

Conclusions: Surprisingly, these results may reflect resistance among adolescent subjects to some enduring effects of drug self-administration, such as reward learning.

2.2 Introduction

Adolescence among humans is a developmental stage associated with vulnerability to drug use and abuse (Laviola et al., 1999, O'Malley and Johnston, 2007). People aged 18 to 20 years report the highest rates of illicit drug use compared to other age groups, and individuals who initiate drug-taking at age 14 or younger are almost five times more likely to classify themselves as drug dependent in adulthood than those who first tried drugs after age 18 (SAMHSA, 2006). However, whether or not these phenomena can be attributed to a biological vulnerability of adolescent development is unclear, especially given that a manifestation of vulnerability to drugs among humans might be adolescent onset of drug use; if so, few vulnerable individuals would be included in any survey group of adult-onset drug users (Shram et al., 2007). Animal models of adolescent drug intake may help to identify the potential role of biological vulnerability in drug abuse.

In rodents, adolescence can be defined as a transitional period sometime between postnatal days (PND) 28 and 60 (Spear and Brake, 1983, Spear, 2000a, Smith, 2003) and shares several key characteristics with primate adolescence (Laviola et al., 1999, Spear, 2000a, Smith, 2003, Crews et al., 2007). As a gold standard measure of drug abuse liability (Schuster and Thompson, 1969, Meisch, 1982), drug self-administration has been used to study adolescent vulnerability to drug reward and reinforcement. Intravenous (i.v.) drug self-administration by adolescent rodents varies across drugs and subject populations (Levin et al., 2003, Belluzzi et al., 2005, Frantz et al., 2007, Kantak et al., 2007, Levin et al., 2007, Shram et al., 2007, Shahbazi et al., 2008, Doherty et al., 2009), and thus requires further investigation. The first aim of the present study was to replicate and extend our previous results that no age differences exist in cocaine self-

Wistar rats to acquire lever-pressing maintained by i.v. infusions of cocaine. We also compared outcomes using two different drug dosing methods, changing either volume or concentration of the i.v. drug infusions to account for daily changes in body weight.

Long-lasting vulnerability to relapse to drug-seeking, or drug-taking during periods of abstinence, is one of the major challenges for treatment of drug addiction (Hunt et al., 1971, Sayette et al., 2000, Mezinskis et al., 2001, Chung and Maisto, 2006). Yet relapse or

reinstatement of drug-seeking after adolescent drug use remains almost entirely unexplored in animal models. To the best of our knowledge, only one study explicitly analyzed extinction and reinstatement in animals that acquired drug self-administration as adolescents (Shram et al., 2007); nicotine-induced reinstatement of nicotine-seeking did not differ across age groups, but Wistar rats that took nicotine as adolescents extinguished their drug-seeking faster than those that took the drug as adults. The second aim of the present study was to compare patterns of extinction and cue- or drug priming-induced reinstatement of cocaine-seeking after a 30-day abstinence period (Grimm et al., 2001, Shaham et al., 2003) in rats that self-administered cocaine during adolescence (adolescent-onset groups) vs. adulthood (adult-onset groups).

Finally, incubation of drug craving refers to time-dependent increases in cue-induced reinstatement of drug-seeking after abstinence from drug self-administration in rodents (Grimm et al., 2001, Lu et al., 2004c). Although incubation may contribute to the persistent vulnerability to relapse observed in human drug addicts (Chung and Maisto, 2006), it has not been examined after adolescent-onset of drug self-administration in rats. Therefore, the third aim of the present study was to compare time-dependent changes in cue-induced reinstatement after 1, 14, 30 or 60 days of abstinence from cocaine self-administration during adolescence vs. adulthood.

Based on previous data, we predicted that neither age nor titration method would affect cocaine self-administration (Frantz et al., 2007, Crombag et al., 2008). However, based on reports of heightened vulnerability to drug dependence after adolescent onset of drug use among humans, we predicted that rats that acquired cocaine self-administration as periadolescents would show higher levels of reinstatement and more robust time-dependent increases in cue-induced reinstatement of cocaine-seeking (incubation of cocaine craving), compared with rats that acquired self-administration as adults.

2.3 Methods Subjects

Male Wistar rats (Charles River Laboratories, Inc., Wilmington, MA) arrived in the laboratory at either postnatal day (PND) 22 or 70-82 and were housed in groups of 2-3 in a humidity (50%) and temperature controlled (22ºC) vivarium on a 12/12 h light/dark cycle (reverse cycle, with lights on at 1900 h). Rats acclimated to these conditions for 6-8 days prior to the start of experiments. Food and water were available ad libitum except during self-

administration sessions. All subjects were observed and/or weighed daily to assess general health and responsiveness to drug exposure. All procedures were conducted in strict adherence to the “Principles of Laboratory Animal Care” and the National Institute of Health Guide for the Care and Use of Laboratory Animals (NRC, 2003).

Drugs

Cocaine hydrogen chloride was obtained from Mallinckrodt Inc. (Hazelwood, MO). For rats receiving variable concentrations to account for differences in body weight, the

concentration of cocaine stock solution in sterile saline was 2.5 mg/ml and was diluted for individual subjects. For the variable volume group, the concentration of stock solution was 1.25

mg/ml. (See below for details on dosing methods.) Methohexital sodium was obtained from Eli Lilly (Indianapolis, IN).

Surgery

The i.v. catheters for drug self-administration were made as previously described (Caine et al., 1993), with minor modifications including a shorter length of tubing inserted into the jugular vein for adolescents (2 cm) compared with adults (4 cm) (Shahbazi et al., 2008).Rats were anesthetized with an isoflurane/oxygen vapor mixture (4-5% for initial anesthetization and 1.5-3% during surgery), and catheter tubing was passed subcutaneously from the animal‟s back to the right jugular vein, inserted into the vein previously punctured with a 25 gauge needle, and tied gently with suture thread. During recovery, rats received approximately 0.2 ml Timentin (Ticarcillin Disodium and Clavulanate Potassium; 100 mg/ml, i.v.) twice daily on the first two days post-surgery, then once daily throughout the experiment. Catheters were also flushed daily with approximately 0.4 ml heparinized saline (100 USP units/1 ml). Catheter patency was confirmed in all subjects by full loss of muscle tone within 5 s of i.v. infusion of the short acting anesthetic agent, 1% Methohexital sodium, one day before the first and after the last self-

administration session. Subjects that failed either patency test were eliminated from the study. Equipment

Behavioral tests were conducted in operant conditioning chambers enclosed in sound- attenuating, ventilated environmental cubicles (Med Associates, Inc., St. Albans, VT). To start each session, a house light and white noise turned on and two levers extended into the chamber. Lever presses on the inactive lever were recorded but had no scheduled consequences. Presses on the active lever triggered a syringe pump (Med Associates, Inc., St. Albans, VT) to deliver drug solution via a stainless steel swivel (Instech Laboratories, Inc., Plymouth Meeting, PA) and

polyethylene tubing attached to a catheter portal on each animal‟s back. Each reinforced

response lit a cue light above the lever which stayed on for the duration of the infusion. The cue light, house light, and white noise were not present during a 20 s time-out (TO20) after each infusion. Drug delivery and data collection were controlled by Med Associates software (Med PC IV).

Self-Administration

Following a 6-7-day post-surgical recovery, spontaneous acquisition of cocaine self- administration began (PND35 or 83-95), with daily 2 h sessions over 14 days conducted during the dark phase of the light-dark cycle. Non-contingent drug injections were not administered in this phase of experimentation. Lever-pressing on the active lever was reinforced by i.v. infusion of cocaine (0.36 mg/kg/infusion) under a Fixed-Ratio 1 (FR1) TO20 schedule of reinforcement. Infusion Conditions

Two different methods to account for differences in body weight across age groups were compared, in order to verify that a switch from a previous method of adjusting concentration (Frantz et al. 2007) to a preferred current method of adjusting infusion volume would not alter behavioral outcomes. Thus, we tested the influence of varying concentration vs. varying volume of drug infusions on self-administration. Adolescent and adult subjects were assigned to either variable concentration or variable volume conditions. For variable concentration groups, the concentration of cocaine solution for each rat was titrated according to body weight, but the volume was fixed such that all subjects received 0.1 ml cocaine solution over 4 s per infusion (approximately 0.12 ml/kg/s for adolescent, and 0.06 ml/kg/s for adults). For the variable volume groups, the volume of cocaine solution for each rat was titrated according to body weight (0.07 ml/kg/s), but the concentration was fixed such that all subjects received same concentration of

cocaine solution. The volume and duration were based on a standard 0.1 ml/4 s per infusion for a 350 g rat.

Abstinence Period

After 14 days of self-administration, separate groups of rats remained in their home cages for 20-24 h, 14-15, 30-31 or 60-61 days, under normal housing conditions. Rats were handled twice per week during this abstinence period. Cue-induced reinstatement was tested at each abstinence period, but drug-induced reinstatement was tested only after 30-31 days of abstinence, in different experimental groups.

Extinction and Reinstatement

After the abstinence from cocaine self-administration, a within-session extinction and reinstatement test was conducted (Grimm et al., 2001, Grimm et al., 2003). Six 1-h extinction sessions were followed by a 1-h cue- or drug-induced reinstatement test. During extinction, rats were connected to the metal coil tether but not the infusion tubing, white noise remained off, and the house light remained on. Neither cue-lights nor TO signals were presented after presses on either lever. Five-min breaks occurred between each successive session, during which the two levers retracted and the house light turned off.

For cue-induced reinstatement tested after various abstinence periods, rats in the

adolescent-onset groups were PND49-50, 63-64, 79-80, or 109-110, whereas adult-onset groups were PND97-110, 111-124, 127-140 or 157-171. Cue-induced reinstatement tests began with the onset of the house light, white noise, and a 5 s cue-light, followed by a 20 s TO during which the house light, white noise, and cue light were turned off. During the remainder of cue-induced reinstatement sessions, presses on the active lever produced cue sequences identical to those presented during cocaine self-administration, and the pump went on, although no syringe was

loaded. The only difference between self-administration and cue-induced reinstatement sessions was that drug solution was not infused during reinstatement.

For drug priming-induced reinstatement tested 30-31-days after drug self-administration, rats in the adolescent-onset group were PND79-80, whereas the adult-onset group was PND127- 140. Rats were taken out of the chambers after the last extinction session, injected with 10 mg/kg cocaine (i.p.), and placed back into the chamber immediately. This dose was chosen for two reasons. First, 10 mg/kg cocaine (i.p.) is commonly used dose for cocaine priming-induced reinstatement (Schenk and Partridge, 1999, Soria et al., 2008). Second, we tested 10, followed by either 3 or 30 mg/kg cocaine priming-induced reinstatement on successive days in the same animals, and only 10 mg/kg cocaine (i.p.) induced reliable reinstatement. While 3 mg/kg did not induce reinstatement in either age group, 30 mg/kg caused stereotyped behaviors in subjects from each age group. (These preliminary data are confounded by the previous day‟s test and are not shown.) Parameters for drug priming-induced reinstatement sessions were identical to extinction sessions. No control injections of saline were administered before the drug priming- induced reinstatement test.

Data Analysis

For drug self-administration sessions, the number of drug infusions per session was analyzed using a three-way mixed measures analysis of variance (ANOVA) with age and infusion condition (variable concentration vs. variable volume) as between-subjects factors, and day as a within-subjects factor. Total drug intake (mg/kg) summed over the entire 14 days of cocaine self-administration was also compared using a two-way ANOVA, with age and infusion condition as between-subjects factors. The number of lever presses per session was also analyzed using a three-way ANOVA with age as a between-subjects factor, and day and lever (active vs.

inactive) as within-subjects factors. (Data were collapsed across infusion condition for the active vs. inactive levers analysis, because no differences were observed on other measures; see results.)

For extinction sessions, the number of active or inactive lever presses per session was analyzed using a two-way mixed measures ANOVA, with age as a between subjects factor, and session as a within subjects factor. To test for cue- or drug-induced reinstatement of lever- pressing, the number of active lever presses in the last extinction session was compared directly with active lever presses during the reinstatement session, using a two-way age x session ANOVA, with age as a between subjects factor and session (extinction vs. reinstatement) as a within subjects factor. To test our hypothesis regarding age differences in reinstatement directly, the number of active lever presses was also compared across age groups using unpaired t-tests.

For analysis of time-dependent changes in cocaine-seeking, total extinction responses on the active lever summed over all six 1-h extinction sessions were analyzed using a two-way between subjects age x abstinence period ANOVA. Time-dependent changes in the number of active lever presses per reinstatement session were also analyzed using a two-way age x abstinence period ANOVA. Per the analytical methods of Grimm et al. (2001, 2003), one-way ANOVAs were also conducted on each age group separately to test specifically for time-

dependent increases in reinstatement. Unpaired 2-sided t-tests with Bonferroni‟s correction were used for post-hoc comparisons, as appropriate. Results were considered significant if p<0.05.

2.4 Results

Influence of age and infusion condition on cocaine self-administration

All age and titration groups showed similar patterns of reliable cocaine self-

administration, which increased gradually in the first week and stabilized during the second week of testing (n=7-9/group). A three-way age x infusion condition x days ANOVA revealed that

only the main effect of days was significant [F (13, 364)=11.12, p<0.001], but not other main effects nor interactions. Total drug intake (mg/kg) over 14-days of self-administration did not differ across age or infusion condition either (adolescent variable concentration group: 612±50 mg/kg, adult variable concentration group: 704±67 mg/kg, adolescent variable volume group: 666±72 mg/kg, and adult variable volume group: 613±33 mg/kg). Additional groups were added for subsequent extinction and reinstatement tests (n=9-16/group), and all showed similar initial patterns of cocaine self-administration. Thus, we collapsed data across infusion conditions and subsequent abstinence period groups to show similar lever presses and total intake in all

adolescent vs. adult rats (Figure. 2.1a). Moreover, despite relatively high rates of non-reinforced responding on the active lever, rats clearly discriminated between active and inactive levers (Figure. 2.1b) suggesting reliable self-administration. However, no age difference was observed in either total active lever presses or inactive lever presses. A three-way age x days x lever ANOVA revealed the main effects of day [F(13, 1638)=3.24, p<0.001], lever [F(1, 126)=271.07, p<0.001] and day x lever interaction [F(13, 1638)=4.22 , p<0.001], but no other main effects nor interactions.

Extinction and reinstatement after a 30-day abstinence period

No age difference was observed in cocaine self-administration (Figure. 2.1a) or extinction sessions (see below). However, after a 30-day abstinence period, rats in the adolescent-onset group showed lower levels of cue-induced reinstatement than the adult-onset group (Figure. 2.2, n=11 or 12/group), in contrast to similar levels of drug-induced reinstatement (Figure. 2.3, n=6 or 10/group).

During extinction tests before cue-induced reinstatement, both age groups exhibited more active lever pressing than inactive lever pressing in the first 1-h session (Figure. 2.2). Lever-

pressing gradually declined in both age groups by the sixth 1-h extinction session. A two-way age x session ANOVA on extinction responding revealed a significant main effect of session [F(5, 105)=21.704, p<0.001], but neither the main effect of age, nor the age x session interaction was significant. During the cue-induced reinstatement test, re-exposure to drug-associated cues triggered lever-pressing behavior in both age groups. A two-way age x session ANOVA on active lever presses in the last extinction session vs. the reinstatement session revealed main effects of session [F(1, 21)=23.257, p<0.001] and age [F(1, 21)=50.840, p=0.03], but no interaction. Moreover, the adolescent-onset group exhibited less lever pressing than adults during the single cue-induced reinstatement session, confirmed by a targeted t-test [t(21)=2.139, p=0.044]. In contrast, inactive lever presses did not differ between the last extinction session and reinstatement.

During extinction tests before drug-induced instatement, subjects exhibited similar rates of lever-pressing and extinction (Figure. 2.3). Subsequent i.p. injections of cocaine triggered similar levels of reinstatement in both age groups. A two-way age x session ANOVA comparing active lever presses in the last extinction session with the reinstatement session revealed only a main effect of session [F(1, 14)=6.164, p=0.026], but no main effect of age, nor a significant age x session interaction. Inactive lever presses did not differ across age or session.

Time-dependent increases in cue-induced reinstatement

During extinction tests at approximately 1-, 14-, 30- or 60-day abstinence, adolescent- onset groups showed an overall lower level of responding compared with adult-onset groups (Figure. 2.4; n=9-16/group). A two-way ANOVA on the total number of active lever-presses summed over all 6 1-h extinction sessions revealed a main effect of age [F(1,86)=10.30, p=0.002] and abstinence period [F(3, 86)=5.77, p=0.001] but no age x abstinence period interaction.

With regard to cue-induced reinstatement after a 1-day abstinence, both age groups showed similarly low levels of cue-induced lever-pressing (Figure. 2.5). Subsequently, the adult- onset groups demonstrated more robust increases in drug-seeking from 1 to 60 days of

abstinence, compared with the adolescent-onset groups. A two-way ANOVA on active lever- presses showed main effects of age [F(1, 86)=17.29, p<0.001], abstinence period [F(3, 86)=13.82, p<0.001], and an age x abstinence period interaction [F(3, 86)=2.87, p=0.041]. Post-hoc t-tests showed that adolescent-onset groups responded at lower levels than adults after both 30 [F(1, 21)=4.57, p=0.044] and 60 [F(1, 26)=15.61 , p=0.001] days of abstinence. Separate one-way